1. Field of the Invention
The present invention relates to optical systems within the visible and near visible spectrums which include a polarizer apparatus for producing a generally polarized light beam from a generally unpolarized source light beam. More particularly, the present invention relates to such optical systems comprised of a plurality of optical elements, one of which consists of a polarizer having a generally parallel arrangement of elongated elements disposed in the source light beam for interacting with the electromagnetic waves of the source light beam to generally (i) transmit or pass light having a polarization oriented perpendicular to the length of the elements, and (ii) reflect light having a polarization oriented parallel with the length of the elements.
2. Prior Art
Polarized light is utilized in various applications such as, for example, liquid crystal displays (LCDs) and projection systems. Liquid crystal displays are commonly used for displays in laptop computers and other information displaying devices such as watches and calculators. Liquid crystal projectors are also used to display information, but project the information or images onto a distant screen. Such projectors usually have their own, powerful light source.
The liquid crystal display devices within these projectors employ polarizer devices in combination with the properties of the liquid crystal elements to selectively transmit or absorb light to produce a pattern of light and dark pixels, creating the desired image. The ability to turn light on or off leads to their common designation as a liquid crystal light valve. They function by taking advantage of the liquid crystal material's ability to rotate the polarization of light when organized and aligned appropriately, and its characteristic that this proper alignment can be altered by an external electric field.
Typically, two polarizer devices are employed, one on each side of the liquid crystal elements, creating a light valve assembly. The purposes of the polarizer devices are to present polarized light to the liquid crystal elements and then to analyze the light passed by the liquid crystal elements and block light of the undesired polarization.
It should be understood that the first polarizer device that presents light to the liquid crystal elements need not be immediately adjacent to the liquid crystal elements. However, it is required that the light arriving at the liquid crystal elements be highly plane polarized in order to present a quality, high-contrast image. Therefore, a polarized light beam generated by a polarizer device some distance from the liquid crystal elements could function as this first polarizer device. Of course, there are other applications for polarized light beams, such as are found in scientific instruments and certain types of illumination systems.
The term "polarized" or "polarized light" refers to a beam of light generally having a single linear, or planar, polarization defined by similarly oriented electromagnetic waves. A natural beam of light, on the other hand, is generally unpolarized, or has a number of planes of polarization defined by the electromagnetic waves emitted by the light source. This natural, or unpolarized, light may be characterized as being composed of two, orthogonal, linear (plane) polarizations.
The electromagnetic waves of a particular polarization, or orientation, may be separated out from the unpolarized source, which contains both the particular polarization and the orthogonal polarization. Devices that separate out a particular polarization are called polarizers and may be used to obtain a beam of light generally having a single polarization, or linearly polarized light.
The concepts of polarized light and certain polarizing devices have existed for over a century. Surprisingly, the most modern and advanced applications of polarized light still employ polarizers that are fundamentally unchanged from those of over 30 years ago. This situation is surprising because the fundamental physical mechanisms by which these polarizers function do not provide ideal polarizers for most applications. The resulting performance limitations seriously constrain optical system design flexibility, optical efficiency, system cost, and over-all performance. The consequences of these limitations have led to numerous attempts to improve polarizer performance in ways that typically compromise performance in one or more characteristics, in order to obtain less restrictive performance in another characteristic.
Examination of the history of polarizers and their use in optical systems to produce polarized light beams shows that the polarizer component is the primary and most significant reason why the use of polarized light beams exhibits one or more of the following characteristics: inefficiency, color-dependent performance variations, a requirement for highly collimated light, and complicated optical systems.
Probably the first polarizer known was a birefringent polarizer formed from a calcite crystal. Birefringent polarizers can now be made from many crystals and also certain stretched polymers. Birefringent polarizers are formed from materials that have a different optical index in one direction compared to another, though the degree of difference in the optical index will vary with the color of the light. This differing optical index can be used to separate beams of one linear polarization from another, though this separation typically consists of a small angular deviation. This narrow separation may require the use of complicated optics. It may also require that the light travel through a significant amount of material or over an extended optical path, leading to a bulky optical element or design. Finally, the narrow separation makes it difficult to use both polarizations, meaning half of the light is usually discarded or wasted through absorption or other means.
Use of a birefringent polarizer is typically characterized by inefficiency, color-dependent performance variations, a requirement for highly collimated light, and complicated optical systems. The bulky optics and extended optical path impose additional performance and design penalties. For these reasons, birefringent polarizers are not commonly used in optical systems such as image projectors.
Another type of polarizer, developed in the 1930s and still the primary polarizer used in laptop computer displays, is the dichroic polarizer. A dichroic polarizer is a polarizer device that absorbs one polarization and passes the other. Many types of dichroic polarizers have been developed, but the most common type consists of a polymer sheet that has been stretched to orient its molecules and then treated with iodine and/or other materials so that the oriented molecules absorb any polarization of one orientation.
The most significant problem with dichroic polarizers is their absorption of light. Typical stretched polymer sheet polarizers absorb essentially all of one polarization and 15% or more of the desired or passed polarization, leading to an inefficient use of light. All polymer polarizers have other problems as well, such as their low tolerance for heat and sensitivity to photon induced chemical changes that cause the material to yellow or become brittle with use and age. These problems become increasingly critical as the brightness of the optical system is increased. The inherent inefficiency of all dichroic polarizers combined with the environmental (heat and light) sensitivity of the most common polymer sheet polarizers leaves much to be desired.
Still another fundamental polarizer technology is the thin-film polarizer. It uses the Brewster's effect in which light striking the surface of glass or another medium at the Brewster's angle (near 45 degrees) is converted into two polarized beams, one transmitted and the other reflected. Polarization of light by use of the Brewster's angle can only be accomplished effectively over a very narrow angular range. An example of this type of polarizer is U.S. Pat. No. 2,403,731, issued Jul. 9, 1946, to MacNeille. MacNeille included a number of layers that help broaden the acceptance angle range for this type of polarizer, though the acceptance angle range is still limited to a few degrees in most devices. MacNeille prisms are typically manufactured with the polarizing thin films disposed between the large faces of two wedges, forming a cube with the films diagonally disposed in the cube. Thus, the width and depth of the cube are equivalent.
MacNeille prisms also suffer from color-dependent performance because the Brewster's angle will vary somewhat with color, reducing the effectiveness of the polarizer for broad band visible light applications. A further limitation of MacNeille polarizers and related polarizer devices is that the polarization that is not passed is reflected to the side at right angles to the optical axis of the system. This is an inconvenient location for the unused light and significantly restricts the utility of these types of polarizers. Finally, these types of polarizers become very bulky in order to achieve a reasonable physical aperture because the depth of the MacNeille polarizer increases as the width, or aperture, increases due to the angular disposition of the film in the cube. Alternatively, a less bulky, or less deep, MacNeille polarizer may be formed by numerous, small cubes placed side-by-side to span a wider area, or create a wider physical aperture. This, however, requires a complex assembly that is difficult and costly to manufacture.
Having failed to improve the limitations of these traditional polarizers, efforts were turned in the 1960s to the field of cholesteric polarizers. Rather than directly separating the two linear polarizations as before, cholesteric polarizers create circularly polarized light. Cholesteric polarizers use special materials and chemicals with a molecular structure that interacts with light to produce circular polarization. A cholesteric polarizer will reflect light of one circular polarization and transmit the orthogonal circular polarization.
One problem with cholesteric polarizers is that circularly polarized light is not generally useful and must be converted to linearly polarized light. Another problem with these polarizers is that they do not work well at larger angles, or a broad range of incident angles. Another problem is that cholesteric polarizers are not broadband, or have a limited optical bandwidth. The efforts directed to developing cholesteric polarizers are indicative of the weaknesses in traditional polarizers and the magnitude of the effort invested in seeking improvements.
In the last ten years, a polarizer device has been developed in which stretched polymer sheets are made birefringent. See for example U.S. Pat. No. 5,612,820. These stretched sheets reflect one polarization and pass the other. One problem with this type of polarizer is its low extinction ratio of approximately 15. While useful for some applications, this extinction ratio is not sufficient for imaging applications without a secondary polarizer, and the fundamental physics of this device make it doubtful that this characteristic can be significantly improved. This type of polarizer also suffers from the environmental problems mentioned earlier.
The problems with conventional polarizer devices, some of which have been discussed above, have serious implications in the applications of polarized light. For example, poor light efficiency is undesirable in many applications, such as image projection, where power is expensive or its waste has expensive and undesirable consequences. To begin with, the production of light itself is an inefficient process. The most efficient conversion of electrical energy into light energy occurs in fluorescent lights, which have an efficiency of about 40%. Fluorescent lights, however, are not optically bright sources. Bright sources, such as arc lamps and metal halide lamps, on the other hand, are even less efficient in producing light, having an efficiency under 10%. In addition, bright sources, such as arc lamps, commonly require expensive power regulators. Because of the inefficiencies inherent in simply creating the light energy, it is important to make efficient use of the light produced.
Since unpolarized natural light consists of two orthogonal linear polarizations, the fundamental light polarization process can only provide 50% of the light produced in the desired polarization. Any polarizer that absorbs, or otherwise renders the undesired polarization unusable imposes a significant performance penalty on the optical system. Hence, it is desirable to have a polarizer that enables the undesired polarization to be used in order to increase the energy efficiency.
In addition to the cost of the power to create the light, inefficient polarizers also have other expensive and undesirable consequences. For example, an inefficient polarizer requires that excess light be produced for a given application because so much light is discarded. A more powerful light source generates more heat, weight, and size. Fans are required to dissipate the heat, which also require power, add weight, add size, add expense, add noise and create vibration. Thus, inefficient polarizers lead to systems that are expensive to make and use, heavy, bulky, and noisy. One of the main challenges in any optical design is to make efficient use of the light available, a goal made particularly difficult by traditional polarizers. A useful measure of this efficiency is the luminous efficiency (also termed luminous efficacy), which is the ratio of the power of the light delivered in the image to the power provided to the light source.
There are other important optical limitations besides absorbing or wasting light energy. A critical parameter is the range of incident angles light can take and still interact properly with the polarizer device to be correctly polarized. This property can be discussed or described by a number of terms, such as the numerical aperture, the cone angle, etendue, or optical speed. All of these terms discuss in various contexts in optical design and theory the breadth of angles that an optical component can accept. For the purposes of discussion, we will refer to the acceptance angle, meaning the largest angle from the local normal to the polarizer device that light can have and still be properly, or correctly polarized by the device.
For an image projection system, or other applications of a polarized light beam, a brighter beam is always desirable. The brightness of the polarized light beam is determined by several things. Of course, the first factor is the light source itself. A smaller, more powerful source will provide a brighter beam, all else being equal. The other critical factor is the ability of the optical system to gather light from the source and direct it into the useful beam. For natural, unpolarized light, this ability depends fundamentally on the acceptance angle of the optical system. A system that employs a polarizer with narrow or limited acceptance angles cannot gather as much light from a given divergent source as a system that employs a polarizer with broader or wider acceptance angles. Other advantages of wide acceptance angles include the potential for a more compact optical system, less expensive and lower power light sources since more light can be used from a given source, and other related advantages.
Another critical advantage of wide acceptance angles is that it provides significant freedom in the optical design. For example, a polarizer device with narrow acceptance angles must be placed within the optical system within a limited range of positions and angles relative to the optical axis. The consequences of this limitation can be seen in the case of a MacNeille prism, where recapture of the rejected light must occur off to the side of the optical system. This location is not convenient, it increases the size and width of the entire system, and additional components (that are otherwise unnecessary) are required to redirect the light into the useful polarized beam. All these limitations, of course, affect cost and utility of the optical system. On the other hand, a wider acceptance angle would allow the polarizer device to be placed and positioned such that the rejected light is placed where it is most convenient for the design of the optical system, offering the optical engineer choices and possibilities that have not been available before.
These two important characteristics, non-absorption and large acceptance angle, are mutually exclusive in traditional polarizers. Polarizers with large acceptance angles permit more design flexibility because the polarizer need not be positioned and oriented within a narrow range of acceptable incident angles with respect to the source light. In addition, polarizers with large acceptance angles are able to use more divergent source light. Non-absorbing polarizers, on the other hand, are able to be more efficient because the rejected polarization may be recovered. Both of these characteristics, however, are necessary for a polarizer device that is both efficient and flexible.
Great efforts have been directed towards improving the traditional types of polarizers for better performance in the production of polarized beams of light. This effort is evidenced by the numerous patents filed, mostly for cholesteric type polarizers and MacNeille prism type polarizers. See U.S. Pat. Nos. 5,153,752; 5,200,843; 5,283,600; 5,295,009; 5,357,370; 5,422,756; 5,555,186; 5,570,215; 5,579,138; 5,626,408; and 5,653,520. In most instances, the efficiency of these devices is enhanced by either returning the rejected plane of polarization to the light source or by rotating the plane of polarization and redirecting it. Some systems even separate the two polarizations, generate a portion of the image in each polarization, and carefully combine the final image from each polarization. Common terms used to describe optical systems that recover or make use of both polarizations in the light beam include polarization saving, polarization recovery, and polarization recycling.
However, even though there has been an extraordinary number of patents granted for polarizer devices that implement various schemes for polarization recovery, there are only a few commercial devices implementing it on the market today. These devices are represented by U.S. Pat. No. 5,555,186. The first device introduced uses MacNeille prisms and a careful and complex optical design to deal with the limitations discussed above. Essentially, this device incorporates a number of MacNeille prisms arranged in an array with spaces formed in between the prisms. Light from the source is directed towards the prisms by a multi-lens optical array. One polarization is passed directly through the prisms while the other polarization is directed sideways towards the spaces in between the prisms. Mirrors disposed in these spaces redirect the other polarization again into the useful beam. Waveplates, also disposed in the spaces, rotate the other polarization so that it is the same as the passed polarization. This device suffers from the same problems as those for the MacNeille prisms discussed above, including restricting the other polarization to an orthogonal direction to the optical axis, or out the side of the prism. In addition, the acceptance angle is narrow, restricting the choices for the light source to very small, bright arc lamps that are expensive and cannot be designed for outputs above a few hundred watts. Another problem is that the precise optics can become misaligned, affecting its performance.
This device is illustrative of the extent to which traditional polarizers have been arranged and manipulated in order to improve their efficiency. In addition, this device is illustrative of the design constraints placed on the system due to the limitations of the polarizer. For example, the narrow range of incidence angles accepted by the MacNeille prisms require that the source light be substantially non-divergent and that the source light and polarizer be specifically positioned and oriented with respect to each other. Furthermore, the rejected polarization is directed substantially orthogonal to the optical axis, dictating the location and orientation of other optical elements. In this case, the mirrors are required to be disposed in spaces between the prisms as part of a complex assembly to recapture the rejected polarization.
Another less complicated, and less elegant, method and device for improving the efficiency of the polarizer uses a diffuse reflector. This concept is currently used in flat panel displays, such as lap-top computers. The rejected polarization is reflected backwards onto the diffuse reflector that scatters the light and confuses the polarization. The light is then redirected at the polarizer. Such an approach cannot have more than a 75% efficiency for a single pass. Furthermore, scattering the light is particularly undesirable in some applications, such as image projection.
As discussed above, efforts at achieving an efficient and flexible polarizer have been largely unsuccessful, despite the lengthy history of visible light polarizers. While some alterations have succeeded in improving the efficiency of traditional devices, they are complicated and impose severe design restrictions. Other polarizers are relatively more flexible, but incapable of efficient polarization. None of the polarizers developed thus far have the necessary characteristics for efficiently and flexibly converting a source light beam into a generally polarized light beam. Some of these characteristics or criterion have received limited discussion thus far, but include the following:
The first desired or necessary characteristic is that the polarizer divide the source beam into two beams of orthogonal polarization with very little loss of light. Thus, for example, the polarizer should not absorb, scatter, or misdirect (poorly focus) one or both of the beams. This is a problem for dichroic polarizers, which absorb one polarization, and systems utilizing diffuse reflectors, which scatter one polarization.
Another desired or necessary characteristic is that the polarizer effectively separate one polarization from the other. In other words, the light in each beam must be well polarized. This is referred to as the extinction ratio, which is the ratio of the amount of light of the desired polarization to the amount of the undesired polarization. The criteria for an acceptable extinction ratio varies by the application. For example, current display applications require at least a ratio of 100:1, but this is rapidly increasing to a ratio of 1000:1. Extinction ratios as low as 3:1 may be useful, but further treatment is required. It should be noted, however, that the extinction ratio affects the contrast of liquid crystal displays and projectors, with higher extinction ratios providing better contrast. In any event, it is desirable to achieve a high extinction ratio with the polarizer itself, thus eliminating any need for further treatment. This is a problem for birefringent and cholesteric polarizers.
Another necessary or desired characteristic is that the polarizer be achromatic over the visible spectrum, or for wavelengths generally between 450 and 700 nm. This applies, of course, to both beams. Achromatic performance means that the polarizer performance not be color dependent, or only work for certain colors. Traditional polarizers typically exhibit color-dependent performance variations. This is a particular problem for cholesteric, birefringent, and MacNeille polarizers.
Another necessary or desired characteristic is that the polarizer be optically fast, or gather a large amount of light. This is a direct result of the polarizer acceptance angle. First, the polarizer should have a wide acceptance angle in order to effectively capture any light that reaches the device. Second, the polarizer should be sufficiently large to capture all the incoming light possible. Therefore, the polarizer should have a large acceptance angle and a large physical aperture. It is not very helpful if the polarizer can be made very large in area but has a narrow acceptance angle, nor is it sufficient if the acceptance angle can be large, but making a large polarizer is prohibitive. This is a problem for birefringent, MacNeille, and cholesteric polarizers because the acceptance angles are generally small, and especially for MacNeille polarizers because making them large leads to a very bulky or complex optic.
Another necessary or desired characteristic is that the polarizer should impose few, if any, constraints on the design of the optical system. It should be possible to position the appropriate optical elements as desired for an efficient system, rather than as the polarizer demands. In addition, the polarizer should not restrict the optical characteristics of the projection system or other device. Furthermore, the polarizer should be able to direct the two orthogonal polarizations in any direction chosen, and with any reasonable degree of focusing. This is perhaps the single most troubling criterion for traditional polarizing beamsplitters to meet. All the above-mentioned conventional polarizers place undesirable restrictions on the orientation of incoming and outgoing light.
Another characteristic is that the polarizer be rugged and difficult to damage. This has several aspects. First, the polarizer should be able to tolerate rapid temperature increases and prolonged high temperature exposure, with a tolerance for several hundred degrees Celsius being very desirable. Second, the polarizer should be able to resist damage through vibration. Third, the performance of the polarizer should not change due to changes in environment, such as heat and handling. Finally, the polarizer should not experience physical degradation caused by photochemistry or other degradation mechanisms triggered by the light passing through it.
Another characteristic is that the polarizer be inexpensive and easy to manufacture. This applies both to the polarizer itself and any related optical elements or substrates, whether manufactured as separate parts or in units that perform more than one optical function.
These characteristics and criterion are not inclusive, but list many of the important factors necessary for efficient and flexible polarization. To date, no polarizer has successfully demonstrated all these characteristics.
Another polarizer device, called a wire grid polarizer, has not been described thus far because it generally has not been used in visible light applications. Essentially, a wire grid polarizer is a planar assembly of evenly spaced parallel electrical conductors whose length is much larger than their breadth. Waves with a polarization parallel to the conductors are largely reflected while waves of orthogonal polarization are transmitted, or passed through the grid.
It is not surprising that the wire grid polarizer has not been generally applied in visible light applications. Indeed, the historical development of the wire grid polarizer was focused in radio frequency emissions. For example, the wire grid polarizer was first invented in the 1880s and demonstrated with radio waves. The wire grid was made by wrapping a wire around a pair of separated rods. In the 1920s, the wire grid polarizer began to find practical uses in the infrared field. In the 1940s, the wire grid polarizer began to find uses in the radar field. Today, wire grid polarizers are mainly used in the fields of radar, microwaves, and infrared. The wire grid polarizer has been used in these fields because there are few alternative devices, especially for longer wavelengths, and they are fairly easy to fabricate and use, again with the greatest facility for the longer wavelengths.
In addition to separate development paths, the fields in which the wire grid polarizer and the other conventional polarizers were developed are characterized by different goals, perspectives, needs, and applications. Conventional polarizers, as discussed extensively above, were developed exclusively for the visible light and ultra-violet light fields where scientific instruments have used polarized visible and UV light since 1850.
The visible spectrum is characterized by short wavelengths, ranging between 400 to 700 nm (nanometers). In addition, visible light occupies a very narrow spectral range, covering less than one octave, meaning a very large bandwidth is not critical in visible light polarizers. Because of the uniqueness of visible light, it is used in unique applications. Such applications include imaging and information transfer, illumination, and everything we use our eyes for. Therefore, the emphasis in the visible field has been towards efficiency, brightness, contrast, uniform performance for all colors, and has been adapted for the needs of the human eye.
By contrast, the wire grid polarizer was developed for the infrared, microwave, and radar fields. These fields are characterized by Large wavelengths, between one micron and ten centimeters (1000 nm to 10,000,000 nm). Infrared, microwaves, and radar occupy broad ranges of the spectrum, but, of course, cannot be seen. These wavelengths interact with matter in fundamentally different ways from visible light, and are used very differently in their applications. Therefore, the emphasis in these fields has been different from each other, and from the field of visible light.
Perhaps another reason wire grid polarizers have not been generally used in visible light applications is that there was no perceived need for the wire grid polarizer in such applications. As discussed above, numerous polarizers were already available for visible light applications. Therefore, there was no identifiable reason to develop the wire grid polarizer to make it suitable for use in visible light.
Perhaps the most significant factor discouraging use of wire grid polarizers in visible light applications is the prevailing view that wire grid polarizers are characterized by relatively low extinction ratios. Visible light applications typically require higher extinction ratios than provided by wire grid polarizers. In the visible light field, other polarizer devices with higher extinction ratios were available.
As can be seen, the polarization technologies applied in the long wavelengths characterized by radar, microwaves, and the infrared are distinct in structure and in their physics from the polarizers typically used in the visible spectrum. This situation is a natural result of the separate history of technology development in these fields, the availability of appropriate alternative technologies, and the different goals of those skilled in the arts appropriate to each field. These differences continue to segregate polarizer devices for visible light from the polarizer devices for longer wavelengths even today.
One of the isolated occurrences of a wire grid polarizer within a visible light application is disclosed in U.S. Pat. No. 4,688,897, issued Aug. 25, 1987, to Grinberg et al. Grinberg et al. disclosed a wire grid polarizer in a liquid crystal display to reduce parallax. Essentially, the concept was to use the wire grid polarizer as a mirror to reflect a single polarization. The wire grid polarizer is relatively thin and is chemically compatible with the liquid crystal material. Thus, it can be disposed adjacent the liquid crystal without chemical interference and without a gap between the liquid crystal and the polarizer. Elimination of this gap eliminates parallax in the display.
Much like the traditional polarizers discussed above, one problem with this application of a wire grid polarizer is efficiency. Only the reflected polarization of the light entering the display is used to create an image. The passed polarization of the light passes through the polarizer and is discarded. The purpose of the wire grid polarizer in this application was not to create a polarized light beam in any way, but to solve a specific problem with a particular type of liquid crystal display, i.e. parallax. Furthermore, the wire grid polarizer was merely an improved replacement component in a pre-existing optical system.
Another one of the few uses of a wire grid polarizer in a visible light application is disclosed in U.S. Pat. No. 5,383,053, issued Jan. 17, 1995, to Hegg et al. Hegg et al. disclosed a wire grid polarizer in a virtual image display system to improve the reflection/transmission efficiency over conventional beam splitters. Essentially, the concept was to use the wire grid polarizer as a beam splitter. The system involved first reflecting an image off the beam splitter and then reflecting it back through the beam splitter. Conventional beam splitters were inefficient because less than 50% of incident light was first reflected and less than 50% of the reflected light was then transmitted. In other words, the net efficiency of this system was less than 25%. Hegg et al. disclosed using the wire grid polarizer with a polarized image source because the reflection/transmission efficiency of the wire grid was relatively high. Therefore, although the wire grid polarizer had a low extinction ratio, it could still be used as a high efficiency beam splitter, at least with well polarized light. In addition, the purpose of the wire grid polarizer in this application was not to create a polarized light beam in any way (the light was already polarized), but to solve a problem with a virtual image display, i.e. inefficient beam splitters. Furthermore, the wire grid polarizer was merely used as a replacement component in a pre-existing optical system.
As indicated above by Hegg et al., it has been known that a wire grid could be used as a beam splitter. As another example, U.S. Pat. No. 3,631,288, issued Dec. 28, 1971, to Rogers discloses an automobile headlight for emitting polarized light. The purpose of the headlight is to reduce glare for an oncoming or approaching automobile with a polarizer over its windshield oriented perpendicular to the polarization of the headlight.
The headlight has a light source disposed in an enclosure formed by a parabolic, polished metal, reflecting surface and a parabolic reflective polarizer. Light having a first polarization is transmitted through the polarizer while light having a second, orthogonal polarization is reflected from the polarizer. Rogers discloses that the polarizer may be a multilayer birefringent polarizer or wire grid arrays in glass. The reflected light from the polarizer is reflected back to the metal reflective surface where it is reflected back to the polarizer.
In addition, the metal surface alters the linearly polarized reflected beam to elliptically polarized light. A "small amount" of the elliptically polarized light is light of the first polarization and is transmitted through the polarizer while the rest is again reflected by the polarizer back to the reflective metal surface. This process of continually cycling light back and forth between the polarizer and metal surface with a "small amount" being transmitted through the polarizer with each cycle is repeated "ad infinitum."
One problem with this device is its poor efficiency. The best known reflective metal surface, silver, reflects no more than 98% of the incident light. Thus, as the process repeatedly reflects light back to and from the metal surface, more and more light is lost.
In addition, because the emissive light source is positioned within the reflective enclosure, much of the reflected light is directed back into the emitter. "Light source" is defined herein as the light emitter and the optical elements that gather light from the emitter and form it into a beam.
Similarly to the above patent, U.S. Pat. No. 3,566,099, issued Feb. 23, 1971, to Makas, discloses an automobile headlight for emitting polarized light. The headlight has a light source disposed in an enclosure formed by a parabolic, polished metal, reflecting surface and a reflective polarizer. Makas only discloses that the polarizer may be a diffusion or interference type. A quarter wave plate is disposed in front of the polarizer. Light having a first polarization is transmitted through the polarizer while light having a second, orthogonal polarization is reflected from the polarizer. The wave plate changes the polarization of the reflected light as it passes therethrough between the polarizer and the reflector. Like the prior patent, Makas directs the reflected beam back towards the light source.
Positioning the light source within the reflective enclosure such that the light having the polarization which is reflected from the polarizer must pass back into the light source is not desirable. It is generally accepted when working with a bright light source such as headlamps, or even brighter light sources used in current liquid crystal light valve projector systems, that reflecting light back into the light source where it can encounter and be absorbed by the filament is undesirable. Reflected light which falls on the filament will overheat the filament, leading to premature failure of the light source. It is also a poor method for conserving light energy, because the light energy re-absorbed by the filament is not re-radiated as light of the same wavelength going in the desired direction, but as electromagnetic energy spanning the infra-red to the ultraviolet traveling in all directions. Use of a parabolic reflector with the light source filament at the focal point, which is highly desirable for the production of a collimated beam of light, is especially conducive to this problem because it will direct the majority of light reflected back into the source onto the filament.
Rogers teaches the use of a filament with "loose coils" to try to escape this problem, but this would then leave the effect of enlarging the filament in breadth, reducing the degree of collimation produced by the parabolic reflector, and possibly also reducing the energy-to-light conversion efficiency of the light source. Rogers also speculates that altering the geometric relationship between the filament and the focal point of the parabolic reflector may be desirable for various reasons. Alterations of this nature for the purpose of mitigating against reflected light encountering the filament are also likely to reduce the efficiency of the optical system overall. For these and other reasons it is desirable to specifically avoid reflecting light back into the light source as part of a scheme to increase the polarization efficiency.
For these and other reasons, wire grid polarizers for polarization of visible light have continued to be ignored by optical engineers and device manufacturers. Those skilled in the art of the projection and display fields have continued to search for improvements by pursuing refinements of the traditional visible light polarizers. Although these efforts have resulted in several clever and ingenious variations of conventional polarizer systems, these devices are still constrained by the limitations inherent in the polarizer itself.
Therefore, it would be advantageous to develop a polarizer device and apparatus operable within the visible spectrum and near visible spectrum for physically decoupling two orthogonal polarizations of a source light beam into two polarized beams and selectively directing these beams in substantially any direction. It would also be advantageous to develop such a device capable of being positioned at substantially any incidence angle within an apparatus so that significant design constraints are not imposed on the optical system, but substantial design flexibility is permitted. It would also be advantageous to develop such a device for efficiently producing a generally linearly polarized beam of light from a generally unpolarized light source without wasting substantial portions of the source light and without the need for complex and precise optics. It would also be advantageous to develop such a device with a large acceptance angle capable of accepting relatively divergent light.